[00:00] X-warrior: yes, they need their own control-bucket names I think. [00:04] rick_h_, I don't think so, I just tried to bootstrap to the same region from 2 different machines and the second bootstrap removed the other bootstrap machine [00:06] X-warrior: right, but in your environments.yaml you need to create a second block with a different name make sure the control-bucket field is different [00:07] rick_h_, as far as I know, there is no control-bucket on environments.yaml anymore. At least not in default options. [00:07] but maybe I can add it [00:07] X-warrior: right, but I bet it has a generated value. Check your .juju/environments/XXX.jenv [00:08] yes, there is one in this folder [00:08] and there is acontrol-bucket value [00:08] X-warrior: I know I've got two different ec2 'environments' setup in mine and I've used this. One has a default series of precise and one trusty I use for testing. [00:08] right [00:09] so manually assign that in the second block in your environments.yaml to something else and you can juju bootstrap ec2 and juju bootstrap ec2again [00:09] sweet I will try that [00:09] ty [00:09] :D [00:09] good luck === ericsnow_ is now known as ericsnow [01:18] How can I update a config.yaml file from a charm with an error? I tried with upgrade-charm --force and resolved --retry but it seems that the config.yaml updates are not used [01:25] Fixed it I guess. Sorry [01:50] oh no :( volume-ephemeral-storage has been dropped [02:25] I added a storage relation from storage to postgresql, but it seems to keep the data on the same disk as the machine not on a volume. What am I missing? [03:31] hows it going === thumper is now known as thumper-afk === uru_ is now known as urulama === CyberJacob|Away is now known as CyberJacob === CyberJacob is now known as CyberJacob|Away === jam21 is now known as jam1 === jam21 is now known as jam1 === rogpeppe is now known as rogpeppe1 === rogpeppe1 is now known as rogpeppe === thumper-afk is now known as thumper [12:05] How would I go about deploying cs:precise/foo to a trusty host? [12:49] stub: 2 choices. --to lxc or fork it and deploy it as a trusty charm. [12:58] gnuoy, https://code.launchpad.net/~james-page/charms/trusty/nova-compute/fixup-secgroups/+merge/236874 === beuno_ is now known as beuno [14:10] huh. how do I view annotations on a service from the cli? They don't seem to be showing up in juju status like I'd expect? [14:12] has anyone written an annotate plugin? [14:22] jamespage, vxlan mps: [14:22] https://code.launchpad.net/~gnuoy/charms/trusty/neutron-api/vxlan-support/+merge/236891 [14:22] https://code.launchpad.net/~gnuoy/charms/trusty/neutron-openvswitch/vxlan-support/+merge/236890 [14:22] https://code.launchpad.net/~gnuoy/charms/trusty/quantum-gateway/vxlan-support/+merge/236889 [14:57] tvansteenburgh: http://pastebin.ubuntu.com/8479658/ - as i hack on charmtools i'm adding tests around my create template and i'm getting this spit back at me. looks like we have a missing dependency of yaml in our requirements. [14:57] tvansteenburgh: however, PyYAML is in requirements.txt === kadams54_ is now known as kadams54 [14:58] lazyPower: i don't think it's a missing dep [15:00] lazyPower: try running the charm helpers sync script manually and see if it succeeds [15:01] lazyPower: derp, nvm, i missed part of the paste [15:01] tvansteenburgh: i think its due to the fresh venv [15:01] we're just missing a step somewhere [15:01] whats the diff between juju and vagrant [15:02] Valduare: vagrant is a wrapper that works with different virtualization providers to make interfacing from teh command line and scripting easier [15:02] juju is a system orchestration tool that runs on most major cloud providers, but it also has a 'local' and 'manual' provider to work wherever you need it to work [15:03] Valduare: while juju is service orchestration, vagrant has no concept of whats being deployed inside the virtualized container, so it does well at enlistment/provisioning or kicking off config management tools, it cannot persay, relate services. this is where juju shines as relations, lifecycle management, enlistment and provisioning are all within our wheel house [15:04] so there should be a juju provision for vagrant? [15:04] dosaboy, gnuoy: ok this is ready for review: [15:04] https://code.launchpad.net/~james-page/charms/trusty/hacluster/mix-fixes/+merge/235675 [15:04] https://code.launchpad.net/~james-page/charms/trusty/swift-proxy/network-splits-https/+merge/236903 [15:04] https://code.launchpad.net/~james-page/charms/trusty/nova-cloud-controller/network-splits-https/+merge/236907 [15:04] https://code.launchpad.net/~james-page/charms/trusty/neutron-api/network-splits-https/+merge/236906 [15:04] https://code.launchpad.net/~james-page/charms/trusty/keystone/https-multi-network/+merge/236905 [15:04] https://code.launchpad.net/~james-page/charms/trusty/cinder/https-multi-network/+merge/236904 [15:04] https://code.launchpad.net/~james-page/charms/trusty/glance/network-splits-https/+merge/236908 [15:04] Valduare: actually, we consume vagrant to offer a testing platform to users on mac/windows [15:04] jamespage: busy much? [15:04] lazyPower, just a little [15:05] we have a self imposed feature freeze today on new big features into the openstack charms [15:05] * lazyPower nods [15:05] i just lamenting that's a lot of work to be reviewed [15:05] that one deals with HTTPS+HA with multiple network api endpoint configurations [15:05] * lazyPower hattips [15:05] lazyPower: consume vagrant? what do you mean [15:05] keep up the good work [15:05] ta [15:05] ditto [15:05] Valduare: we use vagrant to provide a 'ready to go' solution for investigating/working with juju [15:06] Valduare: https://juju.ubuntu.com/docs/config-vagrant.html === dames is now known as thedac [15:08] tvansteenburgh: i think i resolved it [15:08] tvansteenburgh: i didnt run python setup.py develop before i kicked off 'make check' [15:08] * lazyPower was being silly apparently [15:10] lazyPower: so I could use vagrant to fire up a vm and setup a config file to get juju to auto provision stuff for me? [15:11] Valduare: what do you mean auto provision? [15:11] i want to be clear that we're talking about the same things :) [15:11] I want the stuff that vagrant markets as being able to have everyone on the team running the same setup vm [15:12] Valduare: You could use vagrant to distribute an identical VM for your team to consume to interact with juju. Through that localized copy of vagrant each developer will have a juju-local environment to use for development/testing. [15:13] once they have that environment, they can create/consume bundles to distribute amongst one another for deployment configurations with juju. For example, developer a creates a mediawiki, haproxy, and mysql bundle - then ships that yaml file to developer b [15:13] developer b loads that bundle into juju and has an identifical deployment as developer a [16:41] dosaboy, gnuoy; hold on my merged for now [16:41] just spotted a problem [16:49] gnuoy, https://code.launchpad.net/~james-page/charm-helpers/get-address-in-network-no-fallback/+merge/236935 [16:54] jamespage: approved, but i can;t merge that one [16:54] good catch tho === urulama is now known as urulama-afk === roadmr is now known as roadmr_afk === mattgrif_ is now known as mattgriffin === scuttle|afk is now known as scuttlemonkey === mattgrif_ is now known as mattgriffin === roadmr_afk is now known as roadmr === CyberJacob|Away is now known as CyberJacob [20:56] kirkland, quickstart now has maas support [20:56] this should make bundles on the OB easier/faster [21:00] dosaboy, gnuoy: OK happy with those branches I pinged about earlier with the new charmhelper fix [21:00] ttfn [21:01] jamespage: just fixed the percona issue [21:01] jamespage: need to run by you tho, seem peer_echo is either misplaced or possibly broken [21:02] dosaboy, ? [21:02] which charm? [21:02] jamespage: https://code.launchpad.net/~cts-engineering/charms/trusty/percona-cluster/ipv6/+merge/235582 [21:03] jamespage: seems to work fine now [21:03] jamespage: i fixed a few things but crucially, removed peer_echo() from the cluster-relation-changed [21:03] as it was screwing up the cluster relation settings [21:03] dosaboy, hmm that's bad [21:04] dosaboy, you'll lose all password data doing that [21:04] hmm [21:04] dosaboy, you probably need to extend the default exclusions for peer echo [21:04] jamespage: yeah that sounds right, i'll take a closer look then [21:13] jcastro: cool, what does that mean? [21:13] one less script on the OB [21:14] I'll show you in brussels, I have some ideas on making these more generic [22:20] thedac, hi ! Do you know if you can twiddle the options juju-deployer is called with for a mojo run > [22:20] ? [22:21] Hey, gnuoy [22:21] gnuoy: IIRC you can set the delay time. Give me a few and I will look closer [22:22] thedac, thanks. [22:27] gnuoy: looks like it is just delay in seconds. Patches are welcome :) [22:28] thedac, I just might do that :) [22:28] thanks for looking, much appreciated [22:29] no problem === CyberJacob is now known as CyberJacob|Away